Child and Adolescent Psychiatry and Mental Health
BioMed Central
Open Access
Research
Chronicity of sleep problems in children with chronic illness: a longitudinal population-based study Børge Sivertsen*1, Mari Hysing2, Irene Elgen3, Kjell Morten Stormark4 and Astri J Lundervold2,4 Address: 1Department of Clinical Psychology, University of Bergen, Bergen, Norway, 2Department of Biological and Medical Psychology, University of Bergen, Norway, 3Department of Pediatrics, Haukeland University Hospital, Bergen, Norway and 4Centre for Child and Adolescent Mental Health, Unifob Health, Bergen, Norway Email: Børge Sivertsen* -
[email protected]; Mari Hysing -
[email protected]; Irene Elgen -
[email protected]; Kjell Morten Stormark -
[email protected]; Astri J Lundervold -
[email protected] * Corresponding author
Published: 27 August 2009 Child and Adolescent Psychiatry and Mental Health 2009, 3:22
doi:10.1186/1753-2000-3-22
Received: 18 June 2009 Accepted: 27 August 2009
This article is available from: http://www.capmh.com/content/3/1/22 © 2009 Sivertsen et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abstract Background: The aim of this study was to examine the chronicity of sleep problems in children with chronic illness, and potential predictors of sleep problems. Methods: Using data from a longitudinal total population study in Norway, The Bergen Child Study, data on sleep problems, chronic illness and potential confounders were assessed at ages 79 and 1113. Results: 295 of 4025 (7.3%) children had a chronic illness, and the prevalence of chronic sleep problems was significantly higher in this group compared to children without chronic illness (6.8% versus 3.6%). Sleep problems at the first wave increased the risk of sleep problems at the second wave, also when adjusting for potential confounders (odds-ratio = 5.41). Hyperactivity and emotional problems were also independent risk factors for later sleep problems. Conclusion: These findings call for increased awareness and development of treatment strategies of sleep problems in children with chronic illness.
Background Sleep problems are among the most common complaints in children, and have been linked to a range of negative consequences, including reduced daytime functioning, academic and cognitive deficits as well as increased risk of emotional and behavioural problems [1,2]. Children with chronic illness are at increased risk for sleep problems, and several cross-sectional studies have found an increased rate of sleep problems in children with specific chronic illnesses, including cerebral palsy [3], epilepsy [4], asthma [5], headaches [6], and migraine [7]. In one of
the few population-based studies assessing sleep problems among children with chronic illness, Hysing et al. [8] found that these children reported more problems falling asleep and had more night-time awakenings compared to their healthy peers. Few longitudinal studies of children in the general population have explored the stability of sleep problems, and with mixed findings. In a Swiss study [9] following children from infancy to 10 years, night-time awakenings were found to be both frequent and persistent over time.
Page 1 of 7 (page number not for citation purposes)
Child and Adolescent Psychiatry and Mental Health 2009, 3:22
In contrast, Gregory et al. [10] found a reduction of sleep problems from early childhood to mid-adolescence, and Laberge et al. [11] found a similar reduction in sleep onset problems in children from 10 to 13 years. However, little is known with regards to the chronicity of sleep problems in children with chronic illness, and to the best of our knowledge, no longitudinal population-based studies have investigated the stability of sleep problems over time in this group of children. The increased rate of sleep problems in children with chronic illness may have several potential pathways, some of them suggesting a higher likelihood of chronicity. For example, chronic illness may affect the sleep physiology and sleep systems in disorders with impaired central nervous system (CNS) functioning. Other factors contributing to a chronic trajectory of sleep problems in chronic illness may include higher rates of upper-airway obstruction and BMI (body mass index), as well as emotional and behavioural disorders, which previously has been linked to sleep problems in children with chronic illness [8]. It is also possible that parental stress related to managing their child's chronic illness might contribute to poor implementation of sleep schedules, and thus sleep problems. Based on the same study population as the study by Hysing et al. [8], the current paper linked two waves of the Bergen Child Study (BCS), assessing all children at two time points (79 and 1113 years of age) in order to explore the chronicity of sleep problems in children with chronic illness. We hypothesized that children with chronic illness would report higher rates of both acute and chronic sleep problems than their peers, and that sleep problems would differ between specific subgroups of chronic illnesses. We expected both sleep problems and behavioural and emotional problems to predict subsequent sleep problems.
Methods Study design and subjects Data stem from the first and second wave of the BCS, carried out in the fall 2002 and spring 2006, respectively. The BCS is a longitudinal total population-based study of children in all public and private schools in the city of Bergen, Norway. The protocol and population of the BCS is described in detail elsewhere [8,12]. In short, in the first wave, the target population was 9430 primary school children aged 7 to 9 years, of which 7007 parents gave their informed consent to participate, yielding a response rate of 74.3%. The second wave was conducted in 2006, and in all 5196 children, now aged 11 to 13 years, participated (response rate: 55.1%). A total of 4025 children participated in both waves. In all, 387 children were reported by their parents to have a chronic illness in the second wave. The 295 (7.3%) children who were identified to have such an illness in both waves were included in the present study.
http://www.capmh.com/content/3/1/22
Instruments Chronic illness (wave 2 only) Chronic illness (CI) was defined the following way: All parents responded to a simple question in wave 2 of the BCS regarding whether or not their child had a chronic illness or a disability. Parents who rated such illness/disability as present went on to categorize it as either (1) asthma, (2) epilepsy, (3) diabetes, (4) mental retardation or (5) other illnesses. Parents who endorsed other illness were asked to specify in their own words what that illness was. Of the 5683 children, 387 (9.6%) were reported to have at least one CI. An experienced paediatrician (IE) categorized the illness in subgroups. In the present study three subgroups of chronic illness were identified and included; somatic illness, neurological illness and asthma. Due to the overlap between children with asthma and allergy/ eczema, the children where the parents only reported allergy/eczema were excluded. Thus, CI was defined as reported by parents and only somatic disorders were included (see Table 1 for all included illnesses). Children reported to have psychiatric disorders (n = 25) and specific learning disabilities (n = 6) on the question about physical illness were included in the non-chronically ill group for statistical analyses. Children with more than one chronic illness were categorized to one illness group in the following order: neurological disorders, asthma and somatic illness. Note that children may have more than one diagnosis. Emotional and behavioural disorder (wave 1 and 2) The Strengths and Difficulties Questionnaire (SDQ) [13,14] is a behavioural screening questionnaire for children aged 416 years comprising 25 items, which can be allocated to five subscales with five items each: (1) emotional symptoms, (2) conduct problems, (3) hyperactivity-inattention problems, (4) peer relationship problems and (5) pro-social behaviour. A total difficulty score is computed by combining the first four subscale scores. Each subscale is scored on a three-point scale; 'not true', 'somewhat true', and 'certainly true', with total subscale scores each ranging from 010, and total difficulties score from 040. The SDQ has been extensively validated in various countries (e.g. in population studies of children and adolescents in Nordic countries) [15-17]. The SDQ was completed by the parents in wave 1, whereas in wave 2 the SDQ was provided also by the children. Sleep problems (wave 1 and 2) Child-reported sleep problems were assessed with one question encompassing difficulties with initiating and/or maintaining sleep (DIMS: "Does your child have problems initiating sleep or have frequent awakenings"), rated on a three-point Likert scale ("completely correct" "partly correct" and "not correct"). A dichotomous variable was used for the purposes of the present study, in which responding either "completely correct" or "partly correct" Page 2 of 7 (page number not for citation purposes)
Child and Adolescent Psychiatry and Mental Health 2009, 3:22
http://www.capmh.com/content/3/1/22
Table 1: Sub-groups of chronic illness in the second wave of the Bergen Child Study*
Subgroups (n) Neurological disorders (76)
Asthma (188) Somatic disorders (55)
Mental retardation and related syndromes Epilepsy Migraine Cerebral palsy Hydrocephalus and myelomeningocele Other
n 27 20 13 6 4 6
Diabetes Gastrointestinal disorders Skeletal disorders Cardiovascular disorders Heamophiliac Kidney Endocrinological disorders Muscle disorders Rheumatism
14 14 12 3 3 3 3 2 1
* Children may have more than one chronic illness
was defined as having DIMS. No data on the time-frame or severity of the sleep problems were available. This operationalization has previously been applied in the BCS [18]. Chronic sleep problems were defined as reporting DIMS at both waves, whereas transient (acute) sleep problems were defined as reporting DIMS at either of the two waves. Demographical/clinical information (wave 2 only) Level of the parental education was reported in three categories (primary school, secondary school and college/ university), while household economy was rated as good, medium or poor by the parents. The child's body mass index (BMI) was calculated as weight (kg) divided by squared height (cm). For the purposes of the present study we used the following percentiles: "underweight": Less than the 5th percentile, "healthy weight": 5th percentile to less than the 85th percentile, "overweight": 85th to less than the 95th percentile, and "obese": Equal to or greater than the 95th percentile [19]. Statistics Pearson Chi-Square Tests and Kruskal-Wallis analysis of variance (ANOVA) with multiple comparisons were used to examine differences on demographics, clinical characteristics and sleep variables, in children with and without chronic illness. Wilcoxon Signed Ranks Test was used to examine differences in the prevalence of sleep problems in the whole sample. Non-parametrical tests were chosen due to the non-normality of the data. Logistic regression analyses were used to further explore the association between chronic illness and sleep problems. In general, logistic regression analysis is considered a robust and appropriate analysis also in non-normal data. Both unadjusted (crude) analyses, as well as separate analyses adjusting for A) gender and age, B) income, education and BMI,
C) parent-reported behavioural problems, and D) childreported behavioural problems were conducted. The rationale for including behavioural problems at both waves in the regression model was to investigate the effect of both previous and co-existing behavioural problems on sleep problems. A fully adjusted analysis including all the listed potential confounders was also conducted. Finally, logistic regression analyses were conducted with the SDQfactors as the exposure variable on subsequent sleep problems. Results are presented as odds ratios (OR) with 95 percent confidence intervals. Analyses were performed using SPSS for Windows 17, and the alpha level was set at a two-tailed 5%. Ethics The study was approved by the National Data Inspectorate and the Regional Committee for Medical and Health Research Ethics in western Norway. Written informed consent was obtained from all parents included in this study. Participants received no payment to participate.
Results Sample characteristics There were significantly more boys than girls in the chronic illness group, a larger proportion was overweight/ obese, and they were more likely to have a lower family income (Table 2). Children with chronic illness also reported significantly higher levels of emotional and behavioural problems at both waves compared to the no chronic illness group. No significant differences were found on age or parental education between the two groups. Chronicity of sleep problems Overall, sleep problems increased significantly during the 4 year-period (8.1% to 12.3%, Z = 7.35, p < .001), with an
Page 3 of 7 (page number not for citation purposes)
Child and Adolescent Psychiatry and Mental Health 2009, 3:22
http://www.capmh.com/content/3/1/22
Table 2: Demographic and clinical characteristics in children with and without chronic illness at wave 2.
Characteristics N Girls, % (n) Wave 1 Age* Emotional and behavioural problems (SDQ Parents-reported)* Emotion Conduct Hyperactivity Peer Total Wave 2 Body-mass index, % (n) Boys Underweight Healthy weight Overweight Obese Girls Underweight Healthy weight Overweight Obese Economy, n (%) Good Medium Poor Education mother, n (%) Primary Secondary College/University Education father, n (%) Primary Secondary College/University Emotional and behavioural problems (SDQ Child-reported)* Emotion Conduct Hyperactivity Peer Total
No chronic illness
Chronic illness
P-value
3730 53.3 (1988)
295 42.7 (126)
< 0.001
8.27 (8.248.30)
8.23 (8.138.33)
.47
1.16 (1.111.21) 0.82 (0.780.86) 2.40(2.332.46) 1.16 (1.111.21) 5.16 (5.025.30)
1.89 (1.652.13) 1.16 (0.991.33) 3.27 (2.983.57) 1.89 (1.652.13) 7.77 (7.038.50)